High-stakes testing may potentially influence young children's attitudes toward school and learning. This study describes the development and evaluation of child and teacher versions of a measure called Children's Attitudes Toward School (CATS). Exploratory factor analyses of responses by 335 first graders and 130 first-grade teachers each identified the same underlying dimensions of attitudes related to three types of school activities--Academic, Early Literacy, and Child-Initiated. Confirmatory factor analyses of the three-factor model among a second split-half sample of 307 children and 129 teachers achieved acceptable fit. In general, children's self-reported attitudes were not associated with their achievement; teachers' judgments of children's attitudes had weak to moderate correlations with children's directly assessed skills and weak to moderately strong correlations with teachers' ratings of children's skills; children's self-reported attitudes and teachers' judgments of children's attitudes were not significantly correlated. Implications of these findings for assessing young children's attitudes are discussed. (Contains 4 tables and 1 figure.)